Object recognition apparatus, and object recognition processing method

Abstract

An object recognition apparatus: identifies a target cluster of reflection points based on a detection target in a range of equal to or greater than a minimum distance that allows for object recognition based on image data of a camera, among measurement data of a ranging sensor; sets a determination region including the target cluster; and recognizes an object based on the image data for the determination region. The object recognition apparatus: extracts clusters of the reflection points included in the measurement data; identifies, from the extracted clusters, the target cluster and a short-distance cluster of the reflection points based on a detection target in a range of less than the minimum distance; recognizes the object based on the image data for the determination region including the target cluster; and recognizes the object based on the reflection points included in the short-distance cluster.

1 . An object recognition apparatus comprising:
 a ranging sensor configured to measure at least distances to reflection points based on reflected waves of applied irradiation waves;   a camera configured to generate image data on an imaging range; and   one or more processors configured to perform an object recognition process based on measurement data of the ranging sensor and the image data of the camera, wherein   the one or more processors are configured to, as the object recognition process,
 identify a target cluster that is estimated to be a group of the reflection points based on a detection target existing in a target distance range, among the measurement data, the target distance range being equal to or greater than a minimum distance that allows for object recognition by the object recognition process based on the image data, 
 set a determination region comprising the target cluster, and 
 recognize an object based on the image data for the determination region, and, 
   in the object recognition process, the one or more processors are configured to
 extract clusters that are groups of the reflection points comprised in the measurement data, 
 identify, from the extracted clusters, the target cluster and a short-distance cluster that is estimated to be a group of the reflection points based on a detection target existing in a distance range of less than the minimum distance that allows for the object recognition, 
 recognize the object based on the image data for the determination region comprising the target cluster, and 
 recognize the object based on the group of the reflection points comprised in the short-distance cluster. 
   
     
     
         2 . The object recognition apparatus according to  claim 1 , wherein
 the one or more processors are configured to set an irradiation energy of each of the irradiation waves to cause the ranging sensor to apply the irradiation wave of a predetermined energy band that allows for detection of the target distance range and the irradiation wave of an energy band other than the predetermined energy band.   
     
     
         3 . An object recognition apparatus comprising:
 a ranging sensor configured to measure at least distances to reflection points based on reflected waves of applied irradiation waves;   a camera configured to generate image data on an imaging range; and   one or more processors configured to perform an object recognition process based on measurement data of the ranging sensor and the image data of the camera, wherein   the one or more processors are configured to, as the object recognition process,
 identify a target cluster that is estimated to be a group of the reflection points based on a detection target existing in a target distance range, among the measurement data, the target distance range being equal to or greater than a minimum distance that allows for object recognition by the object recognition process based on the image data, 
 set a determination region comprising the target cluster, and 
 recognize an object based on the image data for the determination region, and, 
   in the object recognition process, the one or more processors are configured to
 further identify, as a super-resolution target cluster among the measurement data, the target cluster that is estimated to be a group of the reflection points based on a detection target existing in a region, of the target distance range, in which the ranging sensor has a spatial resolution equal to or higher than a spatial resolution of the camera, and 
 perform super-resolution processing on the image data acquired in time series and thereafter recognize the object based on the image data, for the determination region comprising the super-resolution target cluster. 
   
     
     
         4 . An object recognition processing method comprising, as an object recognition process:
 identifying, among measurement data of a ranging sensor configured to measure at least distances to reflection points based on reflected waves of applied irradiation waves, a target cluster that is estimated to be a group of the reflection points based on a detection target existing in a target distance range, the target distance range being equal to or greater than a minimum distance that allows for object recognition by the object recognition process based on image data of a camera configured to generate the image data on an imaging range;   setting a determination region comprising the target cluster; and   recognizing an object based on the image data for the determination region, wherein   the object recognition process comprises
 extracting clusters that are groups of the reflection points comprised in the measurement data, 
 identifying, from the extracted clusters, the target cluster and a short-distance cluster that is estimated to be a group of the reflection points based on a detection target existing in a distance range of less than the minimum distance that allows for the object recognition, 
 recognizing the object based on the image data for the determination region comprising the target cluster, and 
 recognizing the object based on the group of the reflection points comprised in the short-distance cluster. 
   
     
     
         5 . An object recognition processing method comprising, as an object recognition process:
 identifying, among measurement data of a ranging sensor configured to measure at least distances to reflection points based on reflected waves of applied irradiation waves, a target cluster that is estimated to be a group of the reflection points based on a detection target existing in a target distance range, the target distance range being equal to or greater than a minimum distance that allows for object recognition by the object recognition process based on image data of a camera configured to generate the image data on an imaging range;   setting a determination region comprising the target cluster; and   recognizing an object based on the image data for the determination region, wherein   the object recognition process comprises
 further identifying, as a super-resolution target cluster among the measurement data, the target cluster that is estimated to be a group of the reflection points based on a detection target existing in a region, of the target distance range, in which the ranging sensor has a spatial resolution equal to or higher than a spatial resolution of the camera, and 
 performing super-resolution processing on the image data acquired in time series and thereafter recognizing the object based on the image data, for the determination region comprising the super-resolution target cluster.